Study on the practical application of deep learning technologies aimed at achieving low carbon targets in enhancing user experience

Abstract The diverse energy usage behaviors of various types of users within integrated energy systems have heightened the challenges of system coordination and low-carbon operation. To enhance user experience and effectively manage energy consumption, this study, based on an analysis of user behaviors, employs a deep learning architecture that integrates gated recurrent units and convolutional neural networks to classify users precisely and recommend corresponding energy consumption packages. Following rigorous experimentation, the model achieved an accuracy rate exceeding 85% in categorizing users into conservative and aggressive profiles, which significantly enhances user satisfaction.